---

---

1. Introduction

Recently, DeepSeek-R1 was introduced—a lightweight and efficient language model designed for streamlined deployment in resource-constrained environments, such as Node.js. Unlike its larger counterpart, DeepSeek-V3, which utilizes a Mixture of Experts (MoE) architecture with 671 billion parameters (activating only 37 billion per token), DeepSeek-R1 is optimized for speed and simplicity. With 7 billion parameters, it ensures responsive performance while maintaining minimal resource requirements. This adaptation was developed in response to a sudden surge in demand that temporarily overwhelmed the API service, providing a reliable and scalable solution for real-time applications.

2. Brief Description of DeepSeek-R1-7b

DeepSeek-R1-7b is a lightweight language model with 7 billion parameters, optimized for local deployment. Designed for Node.js environments, it ensures low latency and eliminates reliance on cloud services. While retaining core capabilities of its parent model (DeepSeek-V3), such as efficient resource management, DeepSeek-R1-7b features a simplified architecture for rapid deployment on standard hardware. Key advantages include:

• Data Privacy: On-device data processing.
• Low Computational Costs: Optimized for CPU with 8GB RAM.
• Seamless Integration: Pre-built npm packages for quick application integration.

3 How To Run Locally (NodeJS)

1. You need to download and install Ollama.

---

2. Running Ollama on Different Operating Systems

Linux:

Open the terminal: Ctrl + Alt + T or search for "Terminal" in the applications.

macOS:

Open the terminal: Cmd + Space and type "Terminal" or find it in the "Utilities" folder.

Windows:

Open the command prompt: Win + R, type cmd, and press Enter.

---

Paste the command:

ollama run deepseek-r1:7b  

Press Enter. The model will start downloading.

---

3. Download and install NodeJS version 20.16.0.

---

4. Cloning the Repository
   Open the terminal (Linux/macOS) or command prompt (Windows).
   Paste the command to clone the repository:

git clone <repository_link>  

---

5. Running the Project
   Navigate to the directory with the cloned repository:

cd <repository_folder_name>  

---

6. Install Libraries

npm install express cors multer uuid axios  

---

7. Run the Script

On Windows: Run the .bat file. On Linux/macOS

Через команду:

node StreamFlow.js

Через исполняемый файл:

• Создайте файл start.sh

  touch start.sh

• Добавьте в него:

  node StreamFlow.js

• Сделайте его исполняемым:

  chmod +x start.sh

• Запустите:

  ./start.sh

---

8. Usage
   After completing all the steps, the project will be running locally, and you can start using it.

Notes:
Ensure that all necessary dependencies and tools are installed.
If errors occur, check if Ollama and NodeJS are installed correctly and if all steps in the instructions were followed properly.